In Brief [Full Speech Follows]
Subject: ‘Getting Better Value for Money in Public Spending’
Author: Roger Kerr
Date: 13 September 2005
High levels of government spending are a drag on growth, in part because of the deadweight costs of taxation.
Poor quality and badly targeted spending also harm community well-being and growth .
The Business Roundtable supports the inclusion of a taxation and expenditure limit (TEL) rule in the Public Finance Act,
but supplementary disciplines to improve the quality of spending are also needed.
Evidence of wasteful spending abounds. The OECD has noted that around 95 percent of base spending in New Zealand is not
regularly reviewed. This is a damning indictment.
Sources of waste include vested interest lobbying, short-termism in politics, mistaken policy ideas, weaknesses in
public sector management and inappropriate government roles.
Proportional voting systems like MMP tend to increase government spending by about 5 percent of GDP.
There is a need for Treasury guidelines on government spending comparable to the regulatory impact statements that are
required for regulations.
The key requirement should be to establish whether there is a genuine public policy reason for a spending programme. The
main justifications would be the need to ensure the provision of public goods – goods or services that cannot be
supplied by the private sector on a normal commercial basis – and a social safety net.
Existing spending should be reviewed on a regular basis using sound spending guidelines.
Competition for the supply of any private goods and services produced by governments would help ensure value for money.
An independent spending review could have merit, given firm political backing.
CONFERENZ 9TH ANNUAL PUBLIC SECTOR FINANCE FORUM
GETTING BETTER VALUE FOR MONEY IN PUBLIC SPENDING
ROGER KERR
EXECUTIVE DIRECTOR WELLINGTON
NEW ZEALAND BUSINESS ROUNDTABLE 13 SEPTEMBER 2005
GETTING BETTER VALUE FOR MONEY IN PUBLIC SPENDING
My views on government spending can be summarized by the following parable: If you spend your own money on yourself, you
are very concerned about how much is spent and how it is spent. If you spend your own money on someone else, you are
still very much concerned about how much is spent, but somewhat less concerned about how it is spent. If you spend
someone else's money on yourself, you are not too concerned about how much is spent, but you are very concerned about
how it is spent. However, if you spend someone else's money on someone else, you are not very concerned about how much
is spent, or how it is spent.
Milton Friedman
1. Introduction
At this forum last year I presented a paper that summarised a Business Roundtable study, Restraining Leviathan: A Review
of the Fiscal Responsibility Act 1994, by Bryce Wilkinson. It concluded that the FRA (which is now part of the Public
Finance Act) had served New Zealanders well, particularly in constraining budget deficits and debt, and in making the
state of the government’s finances more transparent. However, the legislation has done little to constrain government
spending. A conclusion of the study was that the fiscal responsibility rules should be strengthened by the addition of a
tax and expenditure limit (TEL) provision, along the lines of the fiscal constitutions of a number of US states. This
would be a ‘top down’ constraint that would limit the rate of growth of government spending to population growth plus
inflation, unless voters decided otherwise in a referendum. Surplus revenue would be returned to taxpayers. There would
be provisions to deal with exceptional circumstances.
Fundamental to a TEL rule (which is sometimes called a taxpayer bill rights) is the idea that taxpayers should have more
control over what they are asked to pay in taxes. There was a very revealing exchange on this point earlier this year
when we appeared before parliament’s Finance and Expenditure Committee to present our submission on the 2005 Budget
Policy Statement. The committee’s chairman objected to the idea on the grounds that if taxpayers were asked to authorise
spending increases they would decline to do so. The episode illustrated how detached some in the political establishment
have become from their fundamental role as agents of taxpayers to whom they are accountable.
A TEL would only be one device for improving spending disciplines. It would not, by itself, deal with the problem of
existing spending that is wasteful or poorly targeted. This requires complementary constraints and processes. In this
paper I focus on two important issues – the need for sound guidelines for assessing value for money in spending and the
need for a comprehensive review of base spending. Before doing so I review the nature and scale of the wasteful spending
problem and discuss some of its underlying causes.
2. The cost of wasteful spending
Total government spending currently comprises central government current spending of 30–33 percent of gross domestic
product (GDP), local government current spending of 2–3 percent of GDP, gross capital formation (both sectors combined)
of 2–3 percent of GDP, and other capital spending of perhaps 0.5–1.0 percent of GDP. At close to 40 percent of GDP, this
level of central and local government spending is inconsistent with sustained rapid economic growth. Finance minister
Michael Cullen has demurred and pointed out that four OECD countries with such ratios of government spending achieved
real per capita GDP growth of 4 percent or more since 1985. However, these countries achieved that feat only for periods
of around 5 years. At least a 10-year period is needed to avoid the bias introduced by cyclical factors and to be
consistent with the government’s goal of returning New Zealand to the top half of the OECD income rankings. No
comparable OECD country with a government spending share around New Zealand’s level has achieved per capita GDP growth
of 4 percent or more for a decade. Those that have grown this fast, such as Ireland and South Korea, have lower spending
ratios.
The deadweight costs of taxation alone make it highly unlikely that rapid growth could be achieved with high levels of
government spending, even if it were uniformly of high quality. There is ample evidence, however, that much government
spending in New Zealand is wasteful or poorly targeted. Yet core Crown expenses (excluding contributions to the
Government Superannuation Fund) are currently projected to rise by a further 28 percent, or an additional 2 percent of
GDP, by 2009.
The endemic problem of wasteful spending is evident from routine media reporting. One recent example among many is the
Auditor-General’s report in 2004 on spending by New Zealand Trade and Enterprise and the Visiting Investor Programme. An
editorial in The Dominion Post noted that:
Controls on spending were so loose that it was not always possible to find out who had received money, or how much had
been paid out. Monitoring of outcomes was inadequate … The scenario is becoming depressingly familiar. A government
agency dishes out the money with the best of intent but the worst of control. Inevitably the money is wasted, the
politicians overseeing the spending bluster and there are promises that it will not happen again … [Cabinet minister X]
also resorts to the standard approach adopted when an organisation is found wanting – of saying it is all in the past.
He says he has received assurances that the problems identified by the Audit Office are being addressed.
Unless assistance to businesses is motivated by genuine problems of market failure, most of it is likely to be wasteful
because of its resource misallocation effects.
Another example is the observation by the OECD, in its 2005 survey of New Zealand, that New Zealand-resident enrolments
in post-secondary education have grown by almost 35 percent since 1999, but that this reflected a proliferation of
courses of low quality and/or remote career relevance. This has been common knowledge for many years, but dealing with
it has obviously not been a priority. In the case of health, there has been a 26 percent rise in real spending per
capita since 1999-00 but productivity appears to have fallen. For a range of operations that account for 44 percent of
hospital costs, the Treasury has reported that a recent 7 percent rise in hospital funding produced only a 1.5 percent
increase in outputs.
A good deal of government social spending benefits middle and upper income households, the same groups that
substantially fund it through taxation. A 2001 study found that in 1997/98, 27 percent of spending on health and
education services, retirement incomes and other benefits went to households in the top 40 percent of taxpayers, which
paid around 68 percent of all taxes. Such ‘churning’ of income, and the deadweight costs of taxation associated with it,
does nothing for equity, is economically wasteful and damages economic growth.
The Working for Families package is another example of poor targeting. Around a third of its cost arises from assistance
to families earning over $35,000 a year after tax. It also increases average effective marginal tax rates and thereby
harms growth.
Much wasteful spending occurs in the social welfare area. The United States has demonstrated that the combination of a
freer labour market and stricter conditions of access to benefits can drastically reduce welfare rolls while leaving
most beneficiaries better off. The 35 percent increase since 2000 in the number of adults on sickness and disability
benefits in New Zealand defies logical explanations.
Another indicator of waste is an inability to stop programmes or reprioritise spending. A past secretary to the Treasury
has observed that:
... it was interesting in the 2000 budget round that out of 500 budget proposals from spending ministers, only a
handful offered up savings, particularly given that there was a new government in office, which may have wished to
discontinue or reprioritise some existing programmes.
Although the Labour-led government embarked on a value-for-money review, savings from baseline spending in the 2000
Budget amounted to only around $65 million, 0.2 percent of spending. There has been no sign of ongoing reviews of value
for money in subsequent budgets.
At various times the major government control agencies, the State Services Commission (SSC) and the Treasury, have
commented on government spending programmes. A 1999 SSC Occasional Paper stated that departments were poor at explaining
ex ante how outputs contribute to governments’ desired outcomes. They were also poor at evaluating ex post how well
their outputs contributed to these outcomes. It found that
there were "few incentives for departments to provide robust information on the relationship of their outputs to the
desired outcomes". In its 2002 briefing to the incoming government, the Treasury observed that government agencies could
easily demonstrate efficiency gains in delivering services, but "cannot show to the same extent that services actually
deliver the results that the government wants. This is a serious flaw."
The following statement by the OECD is a damning indictment of New Zealand’s budgetary processes:
There was, and still is, no systematic framework for assessing value for money … Although the budget process forces a
detailed examination of new spending proposals, this represents only 5 percent of total expenditure. There is no
centrally driven, systematic or regular review of the remaining 95 percent.
This is a remarkable statement, especially given New Zealand’s habit of regarding itself as a world leader in public
sector financial management. No company operating in competitive markets could neglect to scrutinise its cost base
regularly. Economic growth and living standards are sacrificed if government consumption spending and transfers do not
represent value for money.
3. What explains excessive and poor quality government spending?
Waste in government can persist on a large scale for many reasons. For example:
(1) vested interests may pressure governments for favours, such as subsidies or under-priced services, at the expense of
taxpayers at large. Current election promises on student loans are a case in point;
(2) the frequently short and uncertain tenure of politicians biases them in favour of policies that provide short-term
political benefits at the expense of much greater future costs (eg national superannuation in 1975);
(3) politicians and state agencies have an interest in protecting state providers from competition, leading to high
costs of production, reduced consumer choice and other inefficiencies in the supply of services (the ACC and Pharmac
monopolies illustrate these points);
(4) the excessive size of parliament and cabinet. Governments have ignored the view of voters in a 1999 referendum that
parliament should be reduced to 100 MPs. A large cabinet means more ministers with spending roles;
(5) cabinet ministers commonly lack expertise in contracting with state agencies and in managing large enterprises.
State agencies may lack expertise in reducing waste;
(6) state ownership of commercial enterprises is typically less efficient than private ownership, and government
provision may crowd out the private sector;
(7) the bureaucratic, rule-bound nature of state welfare programmes make them ill-suited relative to private welfare for
handling the highly personalised nature of many welfare cases;
(8) mistaken policy ideas (such as the drive for self-sufficiency in energy that motivated the ‘Think Big’ programme);
(9) bureaucratic incentives and agendas. State agencies have an incentive to expand their budgets rather than to offer
up savings; and
(10) the sheer size of governments, which exacerbates the problems of bureaucracy and militates against the kind of
scrutiny of costs that occurs in the private sector.
At root these problems are the result of a failure by governments to establish a limited set of priorities and to
execute them with a high degree of competence. The logical end of this road is the loss of any principled concept of the
fundamental purpose of government. This was vividly illustrated when the prime minister was asked why the government was
putting taxpayers' dollars into the America's Cup challenge. She replied: “the Government's role was whatever the
government defined it to be.”
What is needed is more public debate about what role for government will best contribute to the well-being of the
community at large. The OECD and the Treasury in its 2002 briefing to the incoming government took a small step in this
direction by suggesting, inter alia, that governments should not spend taxpayers' money before establishing that the
programme would serve a clearly defined public purpose and was an appropriate role for government.
Unfortunately, these documents left open the question of what is a public purpose or an appropriate role. Standard
economic analysis provides some guidance. It indicates that the first role of government is to protect individual
rights, including security in property and freedom of contract. Other core duties include ensuring the provision, by
state action if necessary, of public goods and a limited social safety net. Where the boundaries around the state are
properly drawn can change through time due to things like changing technologies. There can also be legitimate political
debate over matters such as income redistribution. Nevertheless, limited – not open-ended – government is an essential
underpinning of any society that values freedom and prosperity.
4. What can be done?
In its 2002 survey, the OECD recommended that New Zealand:
- continue to improve the machinery of government (better evaluation, greater focus on outcomes, less fragmentation,
multi-year appropriations);
- raise the effectiveness of social spending (more pro-active administration, greater focus on work incentives);
- make greater use of market mechanisms (contracting-out, competition in education, health and accident insurance);
- strengthen top-down spending control (set timeframes for objectives, fund capital spending in part from
privatisations, maintain fixed departmental baselines); and
- reconsider local government reforms (legislate principles for appropriate activities, establish a top-down control
framework, consider spending caps, encourage efficient user charges, and review the power of general competence).
This is a sensible list although it is not comprehensive. It does not cover issues such as the MMP electoral system, the
number of MPs in parliament, options for making the executive more accountable to parliament, greater use of direct
democracy (strike-down referenda), better spending guidelines, or a spending review commission. Some of these options
were canvassed in section 5 of Restraining Leviathan. More recently, Richard Prebble has put forward some interesting
ideas for making the executive more accountable to parliament.
One of the OECD's most important suggestions is to look for opportunities to make greater use of competition and choice.
Competition is the most powerful mechanism available for improving value for money in respect of private goods. Public
provision of private goods is particularly undesirable because of the conflict of interest it creates when governments
are both regulators of markets and owners of organisations operating in them. This is a problem in energy, water,
transport, accident compensation, education and health.
Time and space do not permit canvassing all the options for obtaining better value for money in public spending. In the
remainder of this paper I explore just two of them, government spending guidelines and the issue of a zero base review
of government spending.
5. Government spending guidelines
Despite Treasury's comment that government agencies lack expertise in evaluating their spending programmes, it is
astonishing that no Treasury guidelines on government spending exist. This contrasts unfavourably with the effort the
Ministry of Economic Development has made to improve the quality of regulatory policy through the requirement in the
Cabinet Office Manual for a Regulatory Impact and Business Compliance Cost Statement to accompany regulatory proposals.
Treasury should encourage a more principled approach to government spending by producing a set of government spending
guidelines and providing training in their use.
What might a good set of guidelines look like? One possibility would be to adopt a framework similar to that adopted in
the guidelines for evaluating government regulations. These require, inter alia, that (1) the problem be defined and its
causes established; (2) a public policy objective be determined that is not so narrow as to prejustify the recommended
policy; (3) all practicable alternatives be identified, including market alternatives, and (4) the benefits from the
recommended course of action should exceed the costs when compared to the best alternative course of action. Such a set
of guidelines for government spending would, if adhered to and applied professionally, make it harder for wasteful
spending to persist.
Nevertheless, experience with regulatory policy making should be heeded. Many poor quality regulatory proposals continue
to be adopted, and many regulatory impact analyses have been weak. In part, this reflects a lack of will and competence
on the part of ministers and departments. It also reflects the limitations of cost benefit analysis. A study
commissioned by the Business Roundtable and other business organisations concluded that additional disciplines of a
constitutional nature should be applied to regulatory policy making.
A similar approach should apply to the evaluation of government spending. Governments should not spend taxpayers’ money
unless:
- for a sound public policy reason (relating to public goods or a social safety net, including underwriting access to
services such as health and education);
- the benefits for citizens are rigorously demonstrated rather than perceived or assumed;
- all realistic alternatives, including voluntary arrangements and non-spending alternatives, have been examined and
found inferior;
- the spending programme conforms with sound constitutional principles, in particular principles governing parliament’s
delegation of the power to tax and spend;
- a careful assessment has been made of the possible unintended and undesired consequences of the programme
(particularly the likely crowding out of voluntary provision, eg private schools or hospitals);
- consideration has been given to whether it will be possible to assess whether the programme is producing the intended
results, and to terminate it if not;
- there is compelling evidence of willingness to pay – ideally evidence of support by a super-majority of taxpayers.
Referenda, as used in Switzerland and other countries, could be used to assess taxpayer willingness to pay for major
programmes; and
- a competent analysis shows that the benefits of spending exceed the costs when compared with the best alternative
course of action.
In considering this checklist it should be noted that unsatisfied wants in the community do not constitute a sound
public policy reason for government spending. Individuals always have unsatisfied wants because they are always
budget-constrained. The government document Opportunities for
All New Zealanders that was released in December 2004 illustrates the importance of this point. It reported on
programmes amounting to 78 percent of core Crown expenses but did not explain problems with private arrangements that
might justify those spending programmes or raise any value-for-money considerations. The implicit assumption was that
whenever there was an unsatisfied need, more government spending was desirable regardless of its efficacy.
Where government programmes are tax-financed, the spending guidelines should explicitly take into account the deadweight
or economic costs of taxation. The current tax system in New Zealand probably imposes economic costs of at least 50
percent for the last dollar of revenue raised. To produce a net gain in welfare, therefore, $1 of marginal government
expenditure has to deliver a benefit of at least $1.50. Many current government programmes would not pass that test.
6. Review of base spending
If a government were determined to undertake a zero base review of government spending, how should it go about the task?
Past experience suggests approaches that would be unlikely to be fruitful. Across-the-board spending cuts and ‘sinking
lid’ policies are arbitrary in their effects in that they may eliminate desirable spending while leaving unjustified
programmes in place.
A more fundamental approach is required. The first step would be a screening aimed at determining which programmes serve
a public purpose and represent an appropriate role for government.
This filter should identify a large number of activities as candidates for opening up to competition or withdrawing
government from entirely.
Leading candidates would be government commercial enterprises. Exiting from these activities both contributes directly
to economic growth and allows governments to focus on performing their core roles better.
The role of governments in providing other private goods should be critically assessed. Greater use of user pays
policies, reduced middle class ‘churning’ and targeted assistance, in combination with private sector competition, would
reveal whether such services represented value for money in the eyes of users.
The remaining programmes would be concentrated in the core public good functions of government that are less amenable to
the disciplines of competition and choice. This necessitates greater emphasis on clarifying objectives, allocating
decision rights and monitoring and rewarding performance. An important task here would be to revitalise public sector
management. The impetus to better performance dating from the second half of the 1980s appears to have waned.
An approach that has been adopted by a number of governments is to set up an independent spending review commission or
task force. Such an exercise has not been undertaken in New Zealand for many years and could have merit. It would allow
the government to tap expertise from outside the public service and get well-informed and independent advice. A review
would desirably be headed by a person with experience in monitoring and control issues in the private and state sectors.
Such a review group should operate on the basis of the spending guidelines discussed in the previous section. It should
assess whether major spending programmes conformed with them and, if not, recommend whether they should be modified or
axed. Because of the scale of the task, the review group should be supported by resources able to make more detailed
investigations where specific expertise is required, for example defence. The Treasury should be fully involved with the
work of such a group. Besides its value to an incumbent government, a high quality report would contribute to
well-informed public debate on public spending, in the same way that past taxation reviews have raised the quality of
debate on tax issues.
A spending review group could also make recommendations on how base spending should be reviewed each year. There seems
no reason why government agencies should not follow best practice in the private sector.
There are many examples of spending commissions and task forces that have produced meagre results. Critical success
factors in such exercises include a focus on core principles for government operations, political determination, and
public understanding of the benefits of more limited government.
It should also be noted that many policy actions that governments can take have indirect impacts on public spending.
Reforms to labour market regulation, for example, can boost job creation and reduce welfare rolls. Tax reductions can
facilitate greater use of charges for government services, at least in the case of better-off groups. Greater use of the
private sector in the provision of infrastructure can reduce government capital expenditure. Regulatory reform can
reduce the need for spending on regulatory agencies.
In general, the adoption of more vigorous pro-growth policies can raise private incomes and reduce demands on
governments for services or assistance.
Reducing the share of government spending in the economy by a combination of keeping spending growth below the growth
rate of the economy and cutting wasteful spending is feasible, as the experience of many countries demonstrates. In its
recent budget the Hong Kong government projects a fall in public spending as a percentage of GDP from 20.2 percent in
2005-06 to 16.0 percent in 2009-10, and a cut in actual spending in that period.
7. Concluding comments
Wasteful spending in government is a worldwide problem that worsened in the industrialised world with the growth in
government from the 1960s. There is evidence that the expansion of government has harmed economic performance without
improving social outcomes.
The will to provide value for money in government spending is weak because accountability to taxpayers and the public
more generally is weak.
Effective responses to this problem require action across a broad front. Fundamental to all of them is the need for much
greater public awareness of the need to focus government on its core activities and to allow competition and choice to
discipline spending wherever possible.
The executive needs to be more accountable to parliament and parliament more responsive to the interests of taxpayers
and less responsive to the lobbying of vested interests that wish to benefit at the expense of the community at large. A
tax and expenditure limitation rule could play an important role in empowering taxpayers.
This paper has focused on two other pieces in the jigsaw, the need for spending guidelines that incorporate a principled
approach to determining a proper role for the state, and the task of reviewing base spending using these guidelines.
Both activities are necessary and desirable; neither will be effective without political determination and progress on
many other fronts.
ENDS